Manifestations of risk-taking behavior in adolescents, such as teen pregnancy and underage drinking, represent major public health problems with substantial economic and social costs. Adolescent risk-taking is multifactorial, and strongly influenced by sociodemographics, gender, peer behavior, and community factors, including neighborhood crime. In addition, evidence suggests that early life exposure to neurotoxicants that impede development of the prefrontal cortex such as polychlorinated biphenyls (PCBs), lead (Pb) and methylmercury (MeHg) may contribute to risk-taking and its neuropsychological correlates. Identifying preventable environmental risk factors that contribute to increased risk-taking behavior is a public health priority, especially among communities exposed to multiple chemical and non-chemical stressors, as is often the case near Superfund sites. The main goals of the proposed research are: (1) to investigate the relation of prenatal exposure to chemical and non-chemical stressors with adolescents' risk-taking and related behaviors, and (2) to combine results from these epidemiologic analyses with contemporary exposure models to build health risk models for the community living near the New Bedford Harbor (NBH) Superfund Site. We will focus our work on prenatal exposure to mixtures of prevalent organochlorines (PCBs, DDE) and neurotoxic metals (Pb, MeHg, manganese, and arsenic), and consider how risks are modified by non-chemical stressors and sociodemographic factors. We will leverage extensive exposure and behavioral measures available from a well-characterized longitudinal birth cohort study ( the New Bedford Cohort or NBC of children born 1993-1998, with 528 followed through adolescence). We will also conduct analyses within a much larger longitudinal population-based dataset of relevant outcomes and sociodemographic risk factors available from the Massachusetts Pregnancy to Early Life Longitudinal (PELL) data system, with over 12,000 New Bedford area births (including all NBC participants) during the same time period. We propose complementary epidemiologic analyses using the NBC and PELL data to characterize the relation of early life chemical and non-chemical exposures with risk-taking behaviors among New Bedford area adolescents born in the 1990s. These analyses will employ state-of-the-art methods for assessing the effects of complex exposure mixtures. The proposed epidemiologic analyses are highly innovative in their capacity to characterize a comprehensive continuum of risk-taking-related outcomes (from psychometric test measures to hospitalizations for clinical disorders). We will combine our epidemiologic models with current measures of multiple metal exposures and novel statistical approaches to characterizing contemporary exposures to multiple chemical and non-chemical stressors, to build robust community health risk models to predict risk-taking behavior and related outcomes for New Bedford. Our health risk estimates will provide insight to the community and other stakeholders about the benefits of interventions that could improve conditions in New Bedford, and thereby decrease future health risk.